About Load Balancing

  • 1 minute(s) read
Prev Next

This document describes the load balancing models supported by Segura, configuration procedures, and recommended best practices. Load balancing is essential to efficiently distribute connections among cluster nodes, ensuring high availability and performance of privileged sessions.

Load balancing models

Segura offers flexibility in load balancing between nodes, supporting:

  • Native load balancing: The cluster automatically distributes incoming connections among available nodes based on internal health and capacity policies.
  • External load balancers: Integration with market-leading solutions such as F5, HAProxy, NGINX, as well as cloud-native load balancers (AWS ELB, Azure Load Balancer, Google Cloud Load Balancer). This supports high-demand scenarios, multi-region environments, and specific architecture requirements.

Configuration procedures

Native load balancing

  1. When adding a new node, Segura automatically includes it in the load balancing pool.
  2. Native health checks verify node availability and capacity before directing connections.
  3. In case of node failure or degradation, new sessions are redirected to healthy nodes.

Integration with external load balancers

  1. Configure each node to accept connections on the designated port (according to supported ports and protocols).
  2. On the load balancer, create a pool/listener pointing to all cluster nodes.
  3. Define the load balancing algorithm (Round Robin, Least Connections, IP Hash, etc.).
  4. Configure health checks to monitor node availability (HTTP, TCP, or custom).
  5. Optionally, configure session affinity/persistence as required by the application flow.

Example topologies

  • On-premises: F5 or HAProxy distribute sessions to nodes in different racks or data centers.
  • Cloud: AWS ELB directs sessions to nodes across multiple availability zones.
  • Multi-region: Local load balancers in each region integrated with a global load balancer for failover and geographic distribution.

Best practices

  • Keep health checks configured for rapid failure detection.
  • Update the load balancer pool when nodes are added or removed.
  • Use load balancers with built-in redundancy and failover for critical environments.
  • Document firewall rules to ensure exclusive communication between the load balancer and nodes in production.